Evren Samur
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Evren Samur has a total of 82 co-authors. Here are the top co-authors by number of shared publications:
Hannes Bleuler
Laboratoire de Systèmes Robotiques, École Polytechnique Féréderale de Lausanne, Switzerland
9 shared publications- Cagatay Basdogan
College of Engineering, Koc University, Istanbul, Turkey
4 shared publications Gholamreza Ilkhani
Department of Mechanical Engineering, Bogazici University, 34342 Bebek, İstanbul, Turkey
4 shared publicationsLionel Flaction
Ecole Polytechnique Fdrale de Lausanne, Lausanne
4 shared publicationsMohammad Aziziaghdam
Department of Mechanical Engineering, Bogazici University, Istanbul, Turkey
4 shared publications- Seref Kemal Talas
Department of Mechanical Engineering, Bogazici University, Istanbul, Turkey
3 shared publications
Latest Publications
- Open Access
- 1 January 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 14, 117389-117405
Raster2Hull: Learning 3D Ship Hull Geometry From Raster 2D General Arrangement and Lines Drawings
- 1 January 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 14, 117389-117405
Open AccessRaster2Hull: Learning 3D Ship Hull Geometry From Raster 2D General Arrangement and Lines Drawings
- 9 October 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Robotics and Automation Letters
- Vol. 10 (11) , 12245-12252
A Soft Reprogrammable Permanent Magnet for Magnetic Soft Robots
Magnetic soft robots promise unprecedented medical applications due to their untethered motion and control capabilities. These magnetic soft robots, which are navigated by external magnetic sources, can be designed to change their shape or mechanical properties under magnetic fields. However, keeping the modified state constant often requires a continuous power supply with high energy consumption. Changing the magnetic properties of soft magnetic robots can overcome this energy demand. To this end, we present a novel soft and reprogrammable permanent magnet whose magnetization can be remotely switched on and off by a short-duration, low- energy external magnetic pulse. It is suitable for on-board use in magnetic soft robots, either to modify the magnetic properties of such robots or to actuate certain functionalities. Based on this concept, two untethered millimeter-scale magnetic soft robots are developed for applications in capsule endoscopy and medical grasping. The predominant feature of the proposed robots is that, apart from navigation, only state change requires an external power supply, and state retention is enabled by the onboard reprogrammable permanent magnet. The proposed novel magnet allows soft magnetic robots to perform specific tasks without physical tethering by enabling remote state change. The presented concept opens new opportunities for magnetic soft robots to achieve energy-efficient, untethered operation with remotely programmable states, offering significant potential for advanced biomedical applications.
- 9 October 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Robotics and Automation Letters
- Vol. 10 (11) , 12245-12252
A Soft Reprogrammable Permanent Magnet for Magnetic Soft Robots
Magnetic soft robots promise unprecedented medical applications due to their untethered motion and control capabilities. These magnetic soft robots, which are navigated by external magnetic sources, can be designed to change their shape or mechanical properties under magnetic fields. However, keeping the modified state constant often requires a continuous power supply with high energy consumption. Changing the magnetic properties of soft magnetic robots can overcome this energy demand. To this end, we present a novel soft and reprogrammable permanent magnet whose magnetization can be remotely switched on and off by a short-duration, low- energy external magnetic pulse. It is suitable for on-board use in magnetic soft robots, either to modify the magnetic properties of such robots or to actuate certain functionalities. Based on this concept, two untethered millimeter-scale magnetic soft robots are developed for applications in capsule endoscopy and medical grasping. The predominant feature of the proposed robots is that, apart from navigation, only state change requires an external power supply, and state retention is enabled by the onboard reprogrammable permanent magnet. The proposed novel magnet allows soft magnetic robots to perform specific tasks without physical tethering by enabling remote state change. The presented concept opens new opportunities for magnetic soft robots to achieve energy-efficient, untethered operation with remotely programmable states, offering significant potential for advanced biomedical applications.
- Open Access
- 8 February 2024
- Published by Springer Nature in Experimental Mechanics
- Vol. 64 (3) , 393-404
Layer Jamming of Magnetorheological Elastomers for Variable Stiffness in Soft Robots
Background: One of the biggest challenges in soft robotics is the variability and controllability of stiffness. Jamming-based approaches have been of interest to change stiffness dramatically by increasing friction between grains, layers, or fibers. Besides, magnetorheological elastomers (MREs) that exhibit magnetic field-dependent viscoelasticity have significant potential as a stiffness variation material. This study investigates the unique mechanics of magnetic jamming of MRE sheets exploring stiffness change both due to jamming and variable viscosity. Methods: Sample MREs and flexible neodymium-iron-boron (NdFeB) magnets are manufactured. Uniaxial tensile tests supported with digital image correlation are performed to characterize the materials. Multi-layer jamming structures comprised of MREs and NdFeB magnets are developed and validated through 3-point bending experiments and finite element simulations. Results: Results show that the stiffness of the multi-layer structure is higher under magnetic field. Furthermore, the stiffness change is increased when MREs are used instead of PDMS as layers. Conclusion: This study proves the concept of magnetic jamming of MRE layers. The results are crucial for the possible soft robotic implementation of the proposed hybrid stiffening approach combining jamming with viscoelasticity modification.
- 8 February 2024
- Published by Springer Nature in Experimental Mechanics
- Vol. 64 (3) , 393-404
Open AccessLayer Jamming of Magnetorheological Elastomers for Variable Stiffness in Soft Robots
Background: One of the biggest challenges in soft robotics is the variability and controllability of stiffness. Jamming-based approaches have been of interest to change stiffness dramatically by increasing friction between grains, layers, or fibers. Besides, magnetorheological elastomers (MREs) that exhibit magnetic field-dependent viscoelasticity have significant potential as a stiffness variation material. This study investigates the unique mechanics of magnetic jamming of MRE sheets exploring stiffness change both due to jamming and variable viscosity. Methods: Sample MREs and flexible neodymium-iron-boron (NdFeB) magnets are manufactured. Uniaxial tensile tests supported with digital image correlation are performed to characterize the materials. Multi-layer jamming structures comprised of MREs and NdFeB magnets are developed and validated through 3-point bending experiments and finite element simulations. Results: Results show that the stiffness of the multi-layer structure is higher under magnetic field. Furthermore, the stiffness change is increased when MREs are used instead of PDMS as layers. Conclusion: This study proves the concept of magnetic jamming of MRE layers. The results are crucial for the possible soft robotic implementation of the proposed hybrid stiffening approach combining jamming with viscoelasticity modification.
- Open Access
A Reinforcement Learning Based Controller to Minimize Forces on the Crutches of a Lower-Limb Exoskeleton
Metabolic energy consumption of a powered lower-limb exoskeleton user mainly comes from the upper body effort since the lower body is considered to be passive. However, the upper body effort of the users is largely ignored in the literature when designing motion controllers. In this work, we use deep reinforcement learning to develop a locomotion controller that minimizes ground reaction forces (GRF) on crutches. The rationale for minimizing GRF is to reduce the upper body effort of the user. Accordingly, we design a model and a learning framework for a human-exoskeleton system with crutches. We formulate a reward function to encourage the forward displacement of a human-exoskeleton system while satisfying the predetermined constraints of a physical robot. We evaluate our new framework using Proximal Policy Optimization, a state-of-the-art deep reinforcement learning (RL) method, on the MuJoCo physics simulator with different hyperparameters and network architectures over multiple trials. We empirically show that our learning model can generate joint torques based on the joint angle, velocities, and the GRF on the feet and crutch tips. The resulting exoskeleton model can directly generate joint torques from states in line with the RL framework. Finally, we empirically show that policy trained using our method can generate a gait with a 35% reduction in GRF with respect to the baseline.
Open AccessA Reinforcement Learning Based Controller to Minimize Forces on the Crutches of a Lower-Limb Exoskeleton
Metabolic energy consumption of a powered lower-limb exoskeleton user mainly comes from the upper body effort since the lower body is considered to be passive. However, the upper body effort of the users is largely ignored in the literature when designing motion controllers. In this work, we use deep reinforcement learning to develop a locomotion controller that minimizes ground reaction forces (GRF) on crutches. The rationale for minimizing GRF is to reduce the upper body effort of the user. Accordingly, we design a model and a learning framework for a human-exoskeleton system with crutches. We formulate a reward function to encourage the forward displacement of a human-exoskeleton system while satisfying the predetermined constraints of a physical robot. We evaluate our new framework using Proximal Policy Optimization, a state-of-the-art deep reinforcement learning (RL) method, on the MuJoCo physics simulator with different hyperparameters and network architectures over multiple trials. We empirically show that our learning model can generate joint torques based on the joint angle, velocities, and the GRF on the feet and crutch tips. The resulting exoskeleton model can directly generate joint torques from states in line with the RL framework. Finally, we empirically show that policy trained using our method can generate a gait with a 35% reduction in GRF with respect to the baseline.
- Open Access
Fiber Jamming of Magnetorheological Elastomers as a Technique for the Stiffening of Soft Robots
There has been a notable focus on the adoption of jamming-based technologies, which involve increasing the friction between grains, layers, or fibers to achieve variable stiffness capability in soft robots. Additionally, magnetorheological elastomers (MREs) that show magnetic-field-dependent viscoelasticity have great potential as a material for varying stiffness. This study proposes a hybrid method (magnetic jamming of MRE fibers) for enhancing the stiffness of soft robots, combining a jamming-based with a viscosity-based method. First, a fiber jamming structure is developed and integrated into a soft robot, the STIFF-FLOP manipulator, to prove the concept of the magnetic jamming of MRE fibers. Then, based on the proposed method, a variable stiffness device actuated by electro-permanent magnets is developed. The device is integrated into the same manipulator and the electronically controlled magnetic jamming and stiffening of the manipulator is demonstrated. The experimental results show that stiffness gain in bending and compression is achieved with the proposed method. The outcomes of this investigation demonstrate that the proposed hybrid stiffening technique presents a promising avenue for realizing variable and controllable stiffness in soft robots.
Open AccessFiber Jamming of Magnetorheological Elastomers as a Technique for the Stiffening of Soft Robots
There has been a notable focus on the adoption of jamming-based technologies, which involve increasing the friction between grains, layers, or fibers to achieve variable stiffness capability in soft robots. Additionally, magnetorheological elastomers (MREs) that show magnetic-field-dependent viscoelasticity have great potential as a material for varying stiffness. This study proposes a hybrid method (magnetic jamming of MRE fibers) for enhancing the stiffness of soft robots, combining a jamming-based with a viscosity-based method. First, a fiber jamming structure is developed and integrated into a soft robot, the STIFF-FLOP manipulator, to prove the concept of the magnetic jamming of MRE fibers. Then, based on the proposed method, a variable stiffness device actuated by electro-permanent magnets is developed. The device is integrated into the same manipulator and the electronically controlled magnetic jamming and stiffening of the manipulator is demonstrated. The experimental results show that stiffness gain in bending and compression is achieved with the proposed method. The outcomes of this investigation demonstrate that the proposed hybrid stiffening technique presents a promising avenue for realizing variable and controllable stiffness in soft robots.
- Open Access
- 17 March 2023
- Published by Royal Society of Chemistry (RSC) in Soft Matter
- Vol. 19 (14) , 2623-2636
Magnetically induced stiffening for soft robotics
Soft robots are well-suited for human-centric applications, but the compliance that gives soft robots this advantage must also be paired with adequate stiffness modulation such that soft robots can achieve more rigidity when needed. For this reason, variable stiffening mechanisms are often a necessary component of soft robot design. Many techniques have been explored to introduce variable stiffness structures into soft robots, such as pneumatically-controlled jamming and thermally-controlled phase change materials. Despite fast response time, jamming methods often require a bulkier pneumatic pressure line which limits portability; and while portable via electronic control, thermally-induced methods require compatibility with high temperatures and often suffer from slow response time. In this paper, we present a magnetically-controlled stiffening approach that combines jamming-based stiffening principles with magnetorheological fluid to create a hybrid mechanical and materials approach. In doing so, we combine the advantages of fast response time from pneumatically-based jamming with the portability of thermally-induced phase change methods. We explore the influence of magnetic field strength on the stiffening of our magnetorheological jamming beam samples in two ways: by exploiting the increase in yield stress of magnetorheological fluid, and by additionally using the clamping force between permanent magnets to further stiffen the samples via a clutch effect. We introduce an analytical model to predict the stiffness of our samples as a function of the magnetic field. Finally, we demonstrate electronic control of the stiffness using electropermanent magnets. In this way, we present an important step towards a new electronically-driven stiffening mechanism for soft robots that interact safely in close contact with humans, such as in wearable devices.
- 17 March 2023
- Published by Royal Society of Chemistry (RSC) in Soft Matter
- Vol. 19 (14) , 2623-2636
Open AccessMagnetically induced stiffening for soft robotics
Soft robots are well-suited for human-centric applications, but the compliance that gives soft robots this advantage must also be paired with adequate stiffness modulation such that soft robots can achieve more rigidity when needed. For this reason, variable stiffening mechanisms are often a necessary component of soft robot design. Many techniques have been explored to introduce variable stiffness structures into soft robots, such as pneumatically-controlled jamming and thermally-controlled phase change materials. Despite fast response time, jamming methods often require a bulkier pneumatic pressure line which limits portability; and while portable via electronic control, thermally-induced methods require compatibility with high temperatures and often suffer from slow response time. In this paper, we present a magnetically-controlled stiffening approach that combines jamming-based stiffening principles with magnetorheological fluid to create a hybrid mechanical and materials approach. In doing so, we combine the advantages of fast response time from pneumatically-based jamming with the portability of thermally-induced phase change methods. We explore the influence of magnetic field strength on the stiffening of our magnetorheological jamming beam samples in two ways: by exploiting the increase in yield stress of magnetorheological fluid, and by additionally using the clamping force between permanent magnets to further stiffen the samples via a clutch effect. We introduce an analytical model to predict the stiffness of our samples as a function of the magnetic field. Finally, we demonstrate electronic control of the stiffness using electropermanent magnets. In this way, we present an important step towards a new electronically-driven stiffening mechanism for soft robots that interact safely in close contact with humans, such as in wearable devices.
- Open Access
Modular Safety-Critical Control of Legged Robots
Safety concerns during the operation of legged robots must be addressed to enable their widespread use. Machine learning-based control methods that use model-based constraints provide promising means to improve robot safety. This study presents a modular safety filter to improve the safety of a legged robot, i.e., reduce the chance of a fall. The prerequisite is the availability of a robot that is capable of locomotion, i.e., a nominal controller exists. During locomotion, terrain properties around the robot are estimated through machine learning which uses a minimal set of proprioceptive signals. A novel deep-learning model utilizing an efficient transformer architecture is used for the terrain estimation. A quadratic program combines the terrain estimations with inverse dynamics and a novel exponential control barrier function constraint to filter and certify nominal control signals. The result is an optimal controller that acts as a filter. The filtered control signal allows safe locomotion of the robot. The resulting approach is generalizable, and could be transferred with low effort to any other legged system.
Open AccessModular Safety-Critical Control of Legged Robots
Safety concerns during the operation of legged robots must be addressed to enable their widespread use. Machine learning-based control methods that use model-based constraints provide promising means to improve robot safety. This study presents a modular safety filter to improve the safety of a legged robot, i.e., reduce the chance of a fall. The prerequisite is the availability of a robot that is capable of locomotion, i.e., a nominal controller exists. During locomotion, terrain properties around the robot are estimated through machine learning which uses a minimal set of proprioceptive signals. A novel deep-learning model utilizing an efficient transformer architecture is used for the terrain estimation. A quadratic program combines the terrain estimations with inverse dynamics and a novel exponential control barrier function constraint to filter and certify nominal control signals. The result is an optimal controller that acts as a filter. The filtered control signal allows safe locomotion of the robot. The resulting approach is generalizable, and could be transferred with low effort to any other legged system.
- 21 March 2022
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 1-6
A Wearable Haptic Guidance System Based on Skin Stretch around the Waist for Visually-Impaired Runners
Visually-impaired people have difficulties to move around in unknown environments. It is even more challenging to do sports such as running. In this study, a wearable haptic guidance system has been developed that can help visually-impaired people to navigate on lanes of any running track individually without requiring external help. The system uses an RGB-D camera and a microcontroller to detect the lanes and to calculate steering angles for navigation. A skin stretch method is used to transfer the steering angles to the user as haptic feedback around the waist. Two user tests were conducted with four participants (one visually-impaired and three able-bodied volunteers) in order to evaluate the proposed system. In the first test, the participants were asked to identify the given haptic feedback on their waist while standing still. In the second test, the system as a whole was evaluated by asking the participants to run the 400 metres on a lane of a standard outdoor running track with the help of the haptic guidance system. The able-bodied participants were blindfolded during the tests. Number of lane violations, number of directional warnings, and total time of run were recorded. Results show that the participants distinguished steering angles with an accuracy of 83.3 %. The participants were also able to complete the 400 metres with reduced lane violations and directional warnings than the reference runs performed without the haptic guidance. Considering these results, we can conclude that the proposed haptic guidance system is a promising tool for visually-impaired individuals who wants to run on a running track alone.
- 21 March 2022
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 1-6
A Wearable Haptic Guidance System Based on Skin Stretch around the Waist for Visually-Impaired Runners
Visually-impaired people have difficulties to move around in unknown environments. It is even more challenging to do sports such as running. In this study, a wearable haptic guidance system has been developed that can help visually-impaired people to navigate on lanes of any running track individually without requiring external help. The system uses an RGB-D camera and a microcontroller to detect the lanes and to calculate steering angles for navigation. A skin stretch method is used to transfer the steering angles to the user as haptic feedback around the waist. Two user tests were conducted with four participants (one visually-impaired and three able-bodied volunteers) in order to evaluate the proposed system. In the first test, the participants were asked to identify the given haptic feedback on their waist while standing still. In the second test, the system as a whole was evaluated by asking the participants to run the 400 metres on a lane of a standard outdoor running track with the help of the haptic guidance system. The able-bodied participants were blindfolded during the tests. Number of lane violations, number of directional warnings, and total time of run were recorded. Results show that the participants distinguished steering angles with an accuracy of 83.3 %. The participants were also able to complete the 400 metres with reduced lane violations and directional warnings than the reference runs performed without the haptic guidance. Considering these results, we can conclude that the proposed haptic guidance system is a promising tool for visually-impaired individuals who wants to run on a running track alone.
- 1 August 2020
- Published by SAGE Publications in Soft Robotics
- Vol. 7 (4) , 521-533
Design and Development of a Growing Pneumatic Soft Robot
Soft continuum robots are getting more popular in areas such as minimally invasive surgery, search and rescue, and inspection due to their inherent compliance and flexibility. However, most of the conventional continuum robots still lack the ability to significantly change size and length. Growth as a means of robotic locomotion is a novel actuation method that can be used to overcome this disadvantage. In this study, we introduce a growing pneumatic soft robot made up of pressurized thin-walled tubings that can move in three-dimensional space with an extension ratio only limited by manufacturing capabilities. Besides the ability to grow from the tip, this design provides active steering by controlling the speed of each tubing separately, controllable stiffness that can be changed during motion, and capability to carry a tool channel. We present models to estimate tip force and position and experimentally verify the force model and robot kinematics. Open-loop speed controller has an overall root mean square error of 2.69% for speeds between 20 and 300 mm/s. The position controller based on the kinematic model has a mean positioning error of 13.9 mm at 100 mm and 22.6 mm at 200 mm longitudinal distance. Robot can produce a tip force of 20.1 N at 150 kPa tubing pressure and reach a maximum speed of 1490 mm/s at 100 kPa. We also demonstrate the navigation capabilities of the robot both in open field and in constrained environments.
- 1 August 2020
- Published by SAGE Publications in Soft Robotics
- Vol. 7 (4) , 521-533
Design and Development of a Growing Pneumatic Soft Robot
Soft continuum robots are getting more popular in areas such as minimally invasive surgery, search and rescue, and inspection due to their inherent compliance and flexibility. However, most of the conventional continuum robots still lack the ability to significantly change size and length. Growth as a means of robotic locomotion is a novel actuation method that can be used to overcome this disadvantage. In this study, we introduce a growing pneumatic soft robot made up of pressurized thin-walled tubings that can move in three-dimensional space with an extension ratio only limited by manufacturing capabilities. Besides the ability to grow from the tip, this design provides active steering by controlling the speed of each tubing separately, controllable stiffness that can be changed during motion, and capability to carry a tool channel. We present models to estimate tip force and position and experimentally verify the force model and robot kinematics. Open-loop speed controller has an overall root mean square error of 2.69% for speeds between 20 and 300 mm/s. The position controller based on the kinematic model has a mean positioning error of 13.9 mm at 100 mm and 22.6 mm at 200 mm longitudinal distance. Robot can produce a tip force of 20.1 N at 150 kPa tubing pressure and reach a maximum speed of 1490 mm/s at 100 kPa. We also demonstrate the navigation capabilities of the robot both in open field and in constrained environments.